majiayu000 Workflow
- Detect: User correction ("No," "Wait," "Wrong," "Not what I meant")
- Trace: Ask "Which instruction allowed this mistake?"
- Project documentation? → Update the source
- Skill workflow? → Edit the skill
- Command logic? → Fix the command
- Agent prompt? → Refine the agent
- Identify the Gap: Why did the instruction fail? Was it missing, vague, or contradictory?
- Patch: Apply targeted edit to the relevant file
- Report: "I've updated [file] to ensure I [new behavior]."
Root Cause Mapping
| If the mistake reveals... | Target... |
|---|---|
| Missing or wrong project rule | Documentation, CLAUDE.md, .claude/rules/ |
| Wrong procedural pattern | Relevant skill in .claude/skills/ |
| Command logic issue | Command file in .claude/commands/ |
| Agent behavior problem | Agent config in .claude/agents/ |
| Hook or tool mismatch | Hook definition or tool schema |
Patching Principles
- Be specific: Patch the exact instruction that failed, not surrounding content
- Generalize: Extract principle from the specific mistake
- Respect structure: Follow each file's existing conventions (XML tags, tables, etc.)
- Place strategically: Put new constraints where they'll be noticed (file bottoms for recency)